Meta-analysis of short-term and long-term survival following repair versus replacement for ischemic mitral regurgitation

Abstract

The optimal surgical strategy for the management of ischemic mitral regurgitation (IMR) is still debated. The purpose of this study was to perform a meta-analysis summarizing the evidence favoring one technique over another (repair vs replacement). A search of the English literature in PubMed was performed using 'ischemic mitral regurgitation' and 'repair or replacement or annuloplasty' in the title/abstract field. Articles were excluded if they lacked a direct comparison of repair versus replacement, or used Teflon/pericardial strip or suture annuloplasty in >10% of the repairs. Nine articles were selected for the final analysis. All studies except one were relatively recent (2004-2009). The patient characteristics between treatment groups were similar across studies. All studies excluded patients with degenerative etiology and used a rigorous definition of IMR. Most patients had concomitant coronary artery bypass graft. In the patients with mitral valve replacement, at least the posterior and, in many cases, the entire subvalvular apparatus were preserved. Mean ejection fraction and proportion of patients with severe ventricular dysfunction were similar between the repair and replacement groups. The odds ratios for the studies, comparing replacement to repair, ranged from 0.884 to 17.241 for short-term mortality and the hazard ratios ranged from 0.677 to 3.205 for long-term mortality. There was a significantly increased likelihood of both short-term mortality (summary odds ratio 2.667 (95% confidence interval (CI) 1.859-3.817)) and long-term mortality (summary hazard ratio 1.352 (95% CI 1.131-1.618)) for the replacement group compared to the repair group. Based on the meta-analysis of the current relevant literature, mitral valve repair for IMR is associated with better short-term and long-term survival compared to mitral valve replacement. Our conclusion should be interpreted in the context of the inherent limitations of a meta-analysis of retrospective studies including heterogeneity of patient characteristics, which may have influenced the physician's decision to perform mitral valve repair or replacement. In the absence of any published randomized studies, mitral procedure selection should be individualized.